Varick Agents

Varick Agents is an applied AI company that designs, builds and deploys custom AI agent systems inside enterprise organizations. Its stated focus is the manual work employees perform between systems — exception routing, cross-system coordination, data reconciliation and other operational tasks the company characterizes as too complex for traditional automation but too repetitive for skilled staff. Rather than selling generalized software, Varick positions its work as end-to-end redesign of departmental workflows, from intake through execution to reporting.

Delivery follows a staged model: an AI opportunity audit mapping how work moves across teams, systems and decision points; process architecture and redesign, in which the agent system is built and integrated with existing enterprise tools, data and operational logic; production deployment into the client's live stack; and ongoing optimization of the agent's decision logic, reliability and scope. The company states that agents are built on top of a client's existing software so no migrations are required, that it can deploy on the client's cloud, and that it will use models from AI labs with which the client holds BAAs.

Target functions include finance operations (accounts payable and receivable, card reconciliation, runway and forecasting, month-end close), procurement (vendor onboarding, purchasing approvals, contract coordination, supplier communication), revenue operations (CRM hygiene, deal desk approvals, pipeline reporting, forecast assembly) and compliance and risk (regulatory deadline tracking, audit documentation, policy monitoring, exception flagging), with the company citing 300+ additional enterprise workflows. A demo site presents function-level demos and three transformation case studies covering finance, sales and operations.

Business model

Varick sells custom AI implementation and deployment engagements to enterprises rather than off-the-shelf software: it audits client operations, architects and builds agent systems integrated with the client's existing tools, deploys them into production (including on the client's own cloud and with the client's contracted model providers), and then continues to optimize and extend the deployed agents. The company describes embedding engineers directly inside client organizations.

Paid enterprise engagements spanning an initial AI opportunity audit, agent system design and build, production deployment and ongoing post-deployment optimization; reported results are tracked as recurring revenue (ARR).

Traction

A third-party report in May 2026 cited public statements by the CEO indicating $220,000 in revenue generated by a two-person team without outside funding, alongside roughly $600 per day of AI API spend. A separate revenue-tracking write-up reported $3M ARR reached in four months. The company itself states that all of its clients reach production deployment and return for further engagements, and its careers page lists numerous open San Francisco roles across engineering, product, sales and operations.

Latest developments

In May 2026 the founder and CEO, Vasuman Moza, publicly contrasted the company's $600 per day Anthropic API spend with a $20 daily employee meal allowance, a post that drew wide commentary about AI-era cost structures and disclosed that the two-person, unfunded company was generating $220,000 in revenue. A later revenue-signal write-up reported the company reaching $3M ARR within four months.